Академический Документы
Профессиональный Документы
Культура Документы
Desalination
journal homepage: www.elsevier.com/locate/desal
University Putra Malaysia, Dept. Civil Engineering, Faculty of Engineering, 43400 UPM Serdang, Selangor, Malaysia
Materials Processing and Technology Lab, Institute of Advanced Technology, 43400 UPM Serdang, Selangor, Malaysia
c
Swinburne University of Technology, Faculty of Engineering and Industrial Sciences, Hawthorn, Melbourne, VIC 3122, Australia
d
Motilal Nehru National Institute of Technology, Dept. Mechanical Engineering, Allahabad-211004, Uttar Pradesh, India
e
Masdar Institute of Science and Technology, Water and Environmental Engineering Program, PO Box 54224, Abu Dhabi, United Arab Emirates
b
H I G H L I G H T S
a r t i c l e
i n f o
Article history:
Received 22 June 2012
Received in revised form 19 November 2012
Accepted 20 November 2012
Available online 21 December 2012
Keywords:
Evaporation
Condensation
Water production
Solar still
Modeling
Desalination
a b s t r a c t
In this paper, a detailed comparison of a few numerical models (with and without considering humid air
properties) for the estimation of water production from a solar water distillation device is investigated. An
extensive laboratory production experiments were executed under fteen sets of external conditions to
nd the properties of evaporation and condensation coefcients to incorporate with the present evaporation
and condensation models (two unique and independent theoretical models), respectively. The calculation accuracy of the evaporation ux computed by two evaporation models (present and previous), Dunkle's and
Ueda's model, and of the hourly condensation ux estimated by two condensation models (present and previous) was examined using the eld experimental results. It was found that the previous evaporation and
condensation models using empirical relationships extremely overestimated and underestimated the observed production ux, respectively. The evaporation ux calculated by the conventional models of Dunkle
and Ueda notably underestimated and overestimated the observed values, respectively. Finally, it is revealed
that the present models have the smallest deviation between the calculated and the observed values among
these six models and can predict the daily production ux.
2012 Elsevier B.V. All rights reserved.
1. Introduction
Energy is a signicant cost in the economics of desalinating waters and conventional fossil fuels have been utilized as the main energy source, but recent concerns over greenhouse gas emissions have
promoted global development and implementation of energy minimization strategies and cleaner energy supplies [1]. Moreover, the rejection of specic constituents, such as boron, in seawater by reverse
osmosis membranes has recently become a concern due to stringent
discharge limits [2]. Therefore, solar stills using renewable (solar) energy would be a potential option to save fossil fuels and to keep the environment clean from hazardous materials.
Corresponding author. Tel.: +60 3 8946 4492; fax: +60 3 8656 7129.
E-mail addresses: ashikcivil@yahoo.com, aahsan@eng.upm.edu.my (A. Ahsan).
0011-9164/$ see front matter 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.desal.2012.11.023
Solar stills of different designs have been investigated by many researchers, e.g. single-slope [3], double-slope basin-type [4], weir-type
cascade [5,6], tubular-type [7], wick-type [8], integrated basin solar
still with a sandy heat reservoir [9], and tilted wick-type with at
plate bottom reector [10]. The performance of solar still can be improved using a hybrid photovoltaic/thermal (PV/T) system [1113],
ash evaporation [14], a hybrid heat pump compression system [15]
and a reciprocating spray feeding system [16]. Eldalil [17] presented
a new concept of active vibratory solar still with an average daily efciency of about 60%. Complicated system is generally costly and may
require regular monitoring with skilled personnel, which makes a
complicated system unsuitable for remote and coastal areas.
To predict the distilled water output, i.e. production, most of the
numerical models of basin-type still have the evaporative mass and
heat transfer correlations using the temperature and vapor pressure
on the water surface and still cover without noting the presence of
174
rectangular black trough inside the tubular cover. The frame can restrain the deformation of a polythene lm (0.15 mm in thickness)
used as a tubular cover. The length and outside diameter of the TSS
were 0.52 and 0.13 m, respectively. The trough is 0.49 m in length
and 0.1 m in width. The solar radiant heat after transmitting through
the cover is mostly absorbed by the saline water in the trough. The
rest is absorbed by the cover and the trough. Thus, the saline water
is heated up and evaporated. The water vapor density of the humid
air increases with the evaporation from the water surface and then
the water vapor is condensed on the inner surface of the cover, releasing its latent heat of evaporation. Finally, the condensed water naturally trickles down toward the bottom of the cover due to gravity
and is stored in a collector.
3. Evaporation and condensation models
3.1. Dunkle's model (for basin-type still)
Dunkle [29] rst proposed a theoretical relation for the convective
heat transfer coefcient, hcw, and a semi-empirical relation for the
evaporative heat ux, qew, given by Eqs. (1) and (2), respectively.
Malik et al. [24] clearly presented these relations (Eqs. (1)(3)) in details using the SI units and noted that the best representation of the
massheat transfer phenomena was obtained if hem / hcw = 0.016273.
Dunkle studied a basin-type solar still but the humid air properties
were not considered in his relations. However, these are still being
used as noted by Murugavel et al. and Dev and Tiwari [18,30], although objections have been raised sporadically about the predictive
accuracy of the fundamental Dunkle's model [31].
"
hcw
#1
3
evw evc T w
0:884 T w T c
268:9 103 evw
3
1
Ag 3
evw evc
Dv
1
Ag 3
evw evha
Dv
where, A 1:
s
me 0:21K o
Islam [33] rst developed two empirical equations for the TSS
without considering the humid air properties. One for evaporation
mass transfer coefcient, hew (m/s), to calculate me and another for
condensation mass transfer coefcient, hcdha (m/s), to calculate the
condensation ux, mc (kg/m 2s), given by Eqs. (6) and (7), respectively. Since these equations do not have theoretical backgrounds,
hcdha 1:55 10
6:50 10
1:97 10
T w T c
T w T c
6
7
evw evha
t
t v K o evw evha
n
zGrSc
q_ s
c l T ha T c
12
0
T c
where, c = 12 and 1 TTha T
: 1 might
ha
c
be affected by the thermal resistance at the
interface and 2 may be inuenced by the
wettability of the cover material. q_ s can
also be expressed using the local heat transfer coefcient of a liquid lm, hl, in the form
q_ s hl T ha T c
13
c l
and ml hl T ha T c =hfg :
"
#1=3
e M v Rv T w T ha vw vha gD2v T w T c
2Rg T ha
10
h 2
i1=3
Rv T w T ha gDv T w T c
where,
hew e Mv2R
:
g T ha
After substituting the physical constants
(Mv, Rv, Rg and g) into Eq. (10), e can be
given as
"
#1=3
0:935me T ha
D2v T w T c
:
T w T ha vw vha
11
Supposedly, e would be inuenced by the
convection (air ow) due to the instability
of the humid air inside the TSS. The temperature difference, Tw Tc, might be one of the
14
15
where, hcdha 0:9963=4
c
me hew vw vha
parameters that represent the degree of instability of the humid air over the water surface. Since Tw is generally higher than Tc, the
humid air may become unstable as the temperature difference Tw Tc (>0) increases.
Based on this idea, a relationship between
e and Tw Tc could be obtained rather
than treated as a constant.
b) Condensation model Ahsan and Fukuhara [35] proposed a modied equation to calculate the apparent heat
ux for the whole area of the tubular cover,
q_ s , as follows:
175
3 3
3
gl l vha l a T ha T a
l Dh3fg vha vc 4
1=4
and
a=Tha Tc /Tha Ta =0.375. After substituting the constants (a and g) into Eq. (15), c
can be formed as
c
4=3
1:24mc
"
#
3
3 1=3
l l vha l T ha T a
:
l Dh3fg
16
Humid air
( R- ) d
Angular
( ) dir.
Liquid
film
( )
Velocity profile
+ d
0.3
e = 0.029+0.0113( Tw-Tc)
8
10
12
14
16
6. Condensation coefcient
Substituting Eq. (17) into Eq. (10), meh (kg/m 2h) is given by
meh
0.1
ds=Rd
Cover
+ d
e 0:029 0:0113T w T c
0.2
"
#1=3
T w T ha vw vha f111:71 43:53T w T c g D2v T w T c
18
T ha
Once the four parameters (Tw, Tha, Tc and RHha) are observed, meh
can be obtained by Eq. (18).
It was observed that RHha varied from 68 to 81% in the steady state
condition during the production experiments under fteen sets of external surroundings, i.e. the presence of dry air in the humid air was
rmly established. Fig. 4 shows a relation between c and ea/eo by
using the data of the above mentioned laboratory production experiments. The c computed by Eq. (16) is inversely proportional to ea/eo
and the regression can be given as
c 0:001710:00112
ea
eo
19
where, eea 1 eevha and evha = f(Tha, RHha). Substituting Eq. (19) into
o
o
Eq. (15), mch (kg/m2h) is dened as
mch
#
"
3
3 1=4
ea 3=4 l l vha l T ha T a
75:7349:6
:
eo
l Dh3fg
20
If the four parameters (Tha, Tc, Ta and RHha) are measured, mch can
be calculated by Eq. (20). Fig. 5 shows the calculation accuracy of meh
calculated by the evaporation model (using Eq. (18)) and of mch calculated by the condensation model (using Eq. (20)), respectively.
The calculated meh or mch and observed mph of laboratory production
experiments are chosen as the vertical and horizontal axes, respectively. It is seen that there is no signicant difference in calculation
accuracy between these two models. Note that the applicability ranges
of these two correlations (i.e. Eqs. (17) and (19)) based on the experimental set-up and observation could be expressed as 1070 C for
[10-4] 15
Condensation coefficient, c
r=R
r=R-
Axial ( x) dir.
dx=1
r
Evaporation coefficient, e
176
12
c =0.00171-0.00112( ea /eo)
9
6
3
0
0.4
0.5
0.6
0.7
0.8
0.9
1.2
Evaporation model
Condensation model
1.0
0.8
0.6
0.4
177
1
8.2. Previous models of TSS
0.2
0.0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
temperatures, 40100% for the relative humidity, and 0.51.5 m for the
length and 0.10.2 m for the diameter of the TSS.
Observed production
Fig. 7(a) and (b) shows the comparison of the calculated evaporation and condensation uxes using the previous models (empirical
Eqs. (6) and (7), respectively) with the observed ones in Fukui and
Muscat, respectively. It is revealed that the previous evaporation and
condensation models extremely overestimated and underestimated
the observed values, respectively.
8.3. Models of Dunkle and Ueda
Fig. 8(a) and (b) shows the comparison of the calculated evaporation ux using the models of Dunkle and Ueda with the observed values
in Fukui and Muscat, respectively. The evaporation ux was computed
by Eq. (3) for Dunkle's model and by Eq. (5) for Ueda's model. It is revealed that the calculated proles considerably differ from the observed
values. The evaporation ux calculated by the models of Dunkle and
Ueda notably underestimated and overestimated the observed values,
respectively. Certainly, the conventional approach cannot calculate the
condensation ux. Hence, it is concluded that the conventional approach
based on the evaporation phenomenon developed for a basin-type still
or for a plate might not be useful to predict precisely the hourly production ux of TSS.
9. Accuracy of present and previous models
Fig. 9(a) and (b) shows the calculation accuracy of the evaporation
ux obtained by two evaporation models (present and previous),
Dunkle's and Ueda's model, and of the hourly condensation ux calculated by two condensation models (present and previous) for the
Calculated evaporation
1.0
1.0
Calculated condensation
0.8
0.6
0.4
0.2
0.8
0.6
0.4
0.2
0
0.0
9 10 11 12 13 14 15 16 17 18 19 20
9 10 11 12 13 14 15 16 17 18 19 20
Fig. 6. Comparison of the hourly calculated evaporation and condensation uxes using the present models with observed production ux. (a) July 5, 2008 in Fukui, Japan and (b) July 13,
2008 in Muscat, Oman.
178
Observed production
1.4
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
Calculated condensation
(by previous model)
Calculated evaporation
(by previous model)
1.2
1.0
0.8
0.6
0.4
0.2
0
9 10 11 12 13 14 15 16 17 18 19 20
9 10 11 12 13 14 15 16 17 18 19 20
Fig. 7. Comparison of hourly calculated evaporation and condensation uxes using previous models with observed production ux. (a) July 5, 2008 in Fukui, Japan and (b) July 13,
2008 in Muscat, Oman.
2
u1 X
t
mphi mehi :
N i1
21
Observed production
Note that meh = mch in Eq. (21) when for condensation would be
calculated. Table 1 shows values for each model. The present evaporation model has the smallest among these six models and for
the present condensation model is higher than that of the present
evaporation model. Furthermore, it is seen that the difference in between the present and previous models is extremely high in both
cases of evaporation and condensation.
The values for both Ueda's and Dunkle's model are much higher
than those for the present models and quantitatively their values
are about 2.5 and 3 times (on average) those of the present models,
respectively. Therefore, it can be concluded that the calculation accuracy of the production ux is certainly improved by the present
models.
A better estimation of the hourly evaporation ux of TSS could be
found using Ueda's model when Ko is replaced with Km (= vKo) and
v is 0.7 in Eq. (5) as follows:
me 0:147K o
Calculated evaporation
(by Dunkle's model)
22
Calculated evaporation
(by Ueda's model)
1.4
1.4
1
Ag 3
evw evha :
Dv
1.2
1.0
0.8
0.6
0.4
0.2
1.2
1.0
0.8
0.6
0.4
0.2
0
0.0
9 10 11 12 13 14 15 16 17 18 19 20
9 10 11 12 13 14 15 16 17 18 19 20
Fig. 8. Comparison of hourly calculated evaporation ux using the models of Dunkle and Ueda with observed production ux. (a) July 5, 2008 in Fukui, Japan and (b) July 13, 2008 in
Muscat, Oman.
1.4
1.4
Evaporation (previous)
Condensation (previous)
Dunkle's model
Evaporation (present)
Condensation (present)
Ueda's model
179
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Fig. 9. Comparison of calculated hourly evaporation ux, meh, and hourly condensation ux, mch, with observed hourly production ux, mph of eld experiments. (a) July 5, 2008 in
Fukui, Japan and (b) July 13, 2008 in Muscat, Oman.
23
using the eld experimental results carried out in Japan and Oman.
The previous evaporation and condensation models extremely
overestimated and underestimated the observed production ux, respectively. The deviations between the present and previous models
were awfully high in both cases of evaporation and condensation. The
evaporation ux calculated by the conventional models of Dunkle and
Ueda notably underestimated and overestimated the observed values,
respectively. The deviations for both Ueda's and Dunkle's model were
about 2.5 and 3 times of the present models, respectively. Finally, it is
revealed that the present models have the smallest deviation among
these six models. It is, therefore, concluded that the present evaporation
and condensation models are the most reliable tool for predicting the
daily production of the Tubular Solar Still.
10. Conclusions
This paper described a detailed comparison of several numerical
models for the estimation of water production from a solar water distillation device. It was observed from the laboratory production experiments that the evaporation coefcient was proportional to the
temperature difference between the water surface and the cover. In
addition, the condensation coefcient was inversely proportional to
the dry air pressure fraction. The calculation accuracy of the evaporation ux obtained by two evaporation models (present and previous),
Dunkle's and Ueda's model, and of the condensation ux calculated
by two condensation models (present and previous) was examined
Nomenclature
a
temperature difference fraction ()
D
diameter of TSS (m)
Dv
molecular diffusion coefcient of water vapor (m 2/s)
e
partial air/vapor pressure in humid air (Pa)
g
gravitational acceleration (9.807 m/s 2)
Gr
Grashof number ()
h
heat/mass transfer coefcient (W/m 2K or m/s)
hfg
latent heat of vaporization (J/kg)
Km
dispersion coefcient of water vapor (kg/msPa)
Ko
diffusion coefcient of water vapor (kg/msPa)
m
evaporation/condensation/production ux (kg/m 2s)
Mv
molecular weight of water vapor (18.016 kg/kmol)
N
number of data points ()
Table 1
Calculation accuracy of the present and previous models.
Root mean squared deviation, (kg/m2/h)
Model
Evaporation
Condensation
Dunkle (meh)
Ueda (meh)
Previous (meh)
Present (meh)
Previous (mch)
Present (mch)
0.34
0.25
0.37
0.09
0.33
0.14
0.30
0.27
0.72
0.08
0.28
0.13
Table 2
Calculation accuracy of the modied models of Dunkle and Ueda.
Model
0.09
0.06
0.09
0.09
180
qew
q_ s
Rg
Rs
Rv
RH
Sc
T
x
v
t
()
Appendix A
Grashof and Schmidt numbers
The local Grashof and Schmidt numbers can be formed [32] as
Gr
3
gx s
2 s
A1
Sc
:
D
A2
Diffusion coefcient
The coefcients, Ko and Dv can be expressed by the following
equations, respectively [32] as
Ko
Dv Mv
Rg T ha
Dv 0:241 10
Subscripts
1
temperature correction coefcient (b1)
2
area fraction (b1) = net area of the liquid lm / whole area
of the tubular cover
a
partial dry air pressure in humid air/ambient air
c
condensation/cover
ch
hourly condensation
cdha
condensation mass transfer coefcient from humid air to
cover
cw
convective heat transfer coefcient from water surface to
cover
e
evaporation
eh
hourly evaporation
em
equivalent mass transfer coefcient from water surface to
cover
ew
evaporative mass transfer coefcient from water surface to
humid air
ha
humid air
l
saturated condensate liquid lm/local coefcient
o
total pressure in humid air (101,325 Pa)
p
production
ph
hourly production
s
saturated/water surface
v
water vapor/vapor pressure
vha
partial pressure of water vapor in humid air
vw
saturated water vapor pressure
w
water
Acknowledgments
The support provided by the University Putra Malaysia, Malaysia
under Research University Grant Scheme (RUGS), 05-05-10-1063RU,
and 9199672 is acknowledged. The authors gratefully acknowledge
Prof. T. Fukuhara, Dr. Shaul Islam, Dr. Saiful Islam, Dr. S. M.
Moniruzzaman, Engr. Keiichi Waki, Dr. Hiroaki Terasaki, Dr. Akihiro
Fujimoto, Dr. Yasuo Kita and Engr. Fumio Asano for their kind cooperation. Special thanks to Dr. Rashid Al Maamari, Dr. Kazuo Okamura and
Mr. Mark Sueyoshi for their continued support. The support provided
by the Ministry of Education, Science, and Culture, Japan; Shimizu Corporation, Japan and Japan Cooperation Center, Petroleum (JCCP), Japan
is also acknowledged.
A3
T ha 1:75
:
288
A4
Although Ko is a function of Tha, the change of Ko in the range of ordinary Tha is small. For example, Ko = 1.93 10 10 kg/msPa for
Tha = 40 C and 2.07 10 10 kg/msPa for Tha = 70 C.
References
[1] A. Subramani, M. Badruzzaman, J. Oppenheimer, J.G. Jacangelo, Energy minimization strategies and renewable energy utilization for desalination: a review, Water
Res. 45 (2011) 19071920.
[2] L.F. Greenlee, D.F. Lawler, B.D. Freeman, B. Marrot, P. Moulin, Reverse osmosis desalination: water sources, technology, and today's challenges, Water Res. 43
(2009) 23172348.
[3] M. Feilizadeh, M. Soltanieh, K. Jafarpur, M.R.K. Estahbanati, A new radiation
model for a single-slope solar still, Desalination 262 (13) (2010) 166173.
[4] K.K. Murugavel, K. Srithar, Performance study on basin type double slope solar
still with different wick materials and minimum mass of water, Renew. Energy
36 (2) (2011) 612620.
[5] F.F. Tabrizi, M. Dashtban, H. Moghaddam, K. Razzaghi, Effect of water ow rate on
internal heat and mass transfer and daily productivity of a weir-type cascade
solar still, Desalination 260 (13) (2010) 239247.
[6] F.F. Tabrizi, M. Dashtban, H. Moghaddam, Experimental investigation of a
weir-type cascade solar still with built-in latent heat thermal energy storage system, Desalination 260 (13) (2010) 248253.
[7] A. Ahsan, M. Imteaz, A. Rahman, B. Yusuf, T. Fukuhara, Design, fabrication and performance analysis of an improved solar still, Desalination 292 (2012) 105112.
[8] J.T. Mahdi, B.E. Smith, A.O. Sharif, An experimental wick-type solar still system:
design and construction, Desalination 267 (23) (2011) 233238.
[9] F.F. Tabrizi, A.Z. Sharak, Experimental study of an integrated basin solar still with
a sandy heat reservoir, Desalination 253 (13) (2010) 195199.
[10] H. Tanaka, Tilted wick solar still with at plate bottom reector, Desalination 273
(23) (2011) 405413.
[11] R. Dev, G.N. Tiwari, Characteristic equation of a hybrid (PV-T) active solar still,
Desalination 254 (13) (2010) 126137.
[12] S. Kumar, G.N. Tiwari, M.K. Gaur, Development of empirical relation to evaluate
the heat transfer coefcients and fractional energy in basin type hybrid (PV/T) active solar still, Desalination 250 (1) (2010) 214221.
[13] S. Kumar, A. Tiwari, Design, fabrication and performance of a hybrid photovoltaic/
thermal (PV/T) active solar still, Energy Convers. Manag. 51 (6) (2010) 12191229.
[14] A.M. El-Zahaby, A.E. Kabeel, A.I. Bakry, S.A. El-agouz, O.M. Hawam, Augmentation of
solar still performance using ash evaporation, Desalination 257 (13) (2010) 5865.
[15] K. Hidouri, R.B. Slama, S. Gabsi, Hybrid solar still by heat pump compression, Desalination 250 (1) (2010) 444449.
[16] A.M. El-Zahaby, A.E. Kabeel, A.I. Bakry, S.A. El-Agouz, O.M. Hawam, Enhancement
of solar still performance using a reciprocating spray feeding systeman experimental approach, Desalination 267 (23) (2011) 209216.
[17] K.M.S. Eldalil, Improving the performance of solar still using vibratory harmonic
effect, Desalination 251 (13) (2010) 311.
[18] R. Dev, G.N. Tiwari, Characteristic equation of the inverted absorber solar still, Desalination 269 (13) (2011) 6777.
[19] N. Setoodeh, R. Rahimi, A. Ameri, Modeling and determination of heat transfer coefcient in a basin solar still using CFD, Desalination 268 (13) (2011) 103110.
[20] J.A. Esfahani, N. Rahbar, M. Lavvaf, Utilization of thermoelectric cooling in a portable active solar stillan experimental study on winter days, Desalination 269
(13) (2011) 198205.
181
[29] R.V. Dunkle, Solar water distillation: the roof type still and a multiple effect diffusion still, in: Proceedings of the International Heat Transfer Conference, ASME,
USA, 1961, pp. 895902.
[30] K.K. Murugavel, S. Sivakumar, J.R. Ahamed, Kn.K.S.K. Chockalingam, K. Srithar,
Single basin double slope solar still with minimum basin depth and energy storing materials, Appl. Energy 87 (2010) 514523.
[31] J.A. Clark, The steady-state performance of a solar still, Sol. Energy 44 (1) (1990)
4349.
[32] M. Ueda, Humidity and Evaporation, in: Corona publishing, Japan, 2000, pp. 83101.
[33] K.M.S. Islam, Heat and vapor transfer in tubular solar still and its production performance. PhD thesis, Department of Architecture and Civil Engineering, University of Fukui, Japan. 2006, pp. 3352.
[34] A. Ahsan, T. Fukuhara, Evaporative mass transfer in tubular solar still, J. Hydrosci.
Hydraul. Eng. 26 (2) (2008) 1525, (JSCE).
[35] A. Ahsan, T. Fukuhara, Condensation mass transfer in unsaturated humid air inside Tubular Solar Still, J. Hydrosci. Hydraul. Eng. 28 (1) (2010) 3142, (JSCE).
[36] P.I. Cooper, Digital simulation of transient solar still processes, Sol. Energy 12
(1969) 313331.
[37] K. Murase, H. Tobataa, M. Ishikawaa, S. Toyama, Experimental and numerical
analysis of a tube-type networked solar still for desert technology, Desalination
190 (2006) 137146.